202 REPORT—1890. 
the volume of the liquid in the cathode vessel increases by 11°09 x37 c.c.,! 
so that the amount of water transferred is 11:09 x 37 x ‘9/18 gramme mole- 
cules. If we trust to the results of the chemical analysis solely to 
identify the ions we must assume that the molecule decomposed is of a 
complex nature, so that the decomposition takes place according to the 
following scheme :— 
Cum(CuSO,)n(H,0); SO, m! (CuS0,) n! (H,0). 
If now we assume that of the molecules decomposed equal numbers are 
taken from the anode and cathode vessels, we can arrange the gains and 
losses as follows :— 
Cathode vessel (1 gramme equivalent of copper deposited). 
Loss. + gramme molecule decomposed. 
1.6.5 +2 (l4+m+m’')Cu 
and + (n+n’)H,O. 
can 2 gramme molecule of the cation. 
0, y Cam (CuSO,) » (Hs O) 
t. eo, 3 + (m+1)Cu 
4 n H,O. 
Hence the net gain of Cu is— 
4 A+m—-m')Cu='75x3Cu. . . 2 @ 
whence 
m— ne! ="5 
and the net gain of water 
ingle \H,0=5— B7EEO.)) Jar ya beellltieel 
n—n' =2'2 x37. 
Hven on this supposition, therefore, the chemical analysis would not 
completely determine the average composition of the molecules decom- 
posed ; only the differences (1—m/’) and (n—w’) are determinable. We 
can, however, assign a formula to the simplest molecule that would give 
the observed result by assuming that the lesser of the two m, m’ and of 
a, v respectively are zero; thus in the case above, assuming that m/=0 
and n’=0, we get m='5 and n=81°4. 
The average molecule in this case would be approximately ? 
Cu, 1(CuS0,), He 0/80, 
or getting rid of fractions 
Cu,(CuS0,)163H,O/(SO,4)o. 
This indicates the extent to which the inferences from chemical analysis 
could be pushed. It is evident on reference to the table given in Wiede- 
mann, 2, p. 592, that the average molecule would be different for different 
degrees of dilution which would alter the numbers on the right-hand side 
of ‘equations (1) and (2). It appears from the table referred to, that if 
1 Wied. Hilec. 2, p. 592. 
2 In a paper in the Proc. Camb. Phil. Soc. (Nov. 1889) the complex molecule is 
erroneously calculated in consequence of my having misunderstood Wiedemann’s 
data, 
